Across the world the media is abuzz with articles excitedly discussing the development of driverless vehicles. Yet a handful of sceptics don’t think it will ever happen, and others think that it will take a long time for the technology to mature, and then for economies to adapt to the widespread production and maintenance of such vehicles. At the same time demand for drivers is high, and is projected to grow in both Europe and the United States. Drivers are also cheap. So it is hard to see what the real incentive is behind the development of driverless.
After 3 years of research into the issue I have concluded that driverless probably will happen, but that it won’t be happening at a significant scale for quite some time. In the meantime drivers are most likely to be impacted by the use of surveillance technology to monitor them and control their work processes.
Drivers probably have 15 years or so to prepare for driverless, although there will be cities and some intercity routes that may see automated vehicles sooner than that. We are already seeing some experiments with driverless vehicles, particularly in the US, and they have already had some collisions and two people, a driver and a pedestrian, have been killed. And we have been seeing the increasing automation of specific driving processes, like parking for example, or braking, over many years. So in this sense the development of driverless tech will be the culmination of a long-term process.
There are a lot of issues to resolve before driverless can be introduced.
One is the issue of choosing a ‘type’ of driverless technology. Driverless vehicles need to be able to communicate with other vehicles, traffic infrastructure and mobile phones. So to develop a driverless vehicle you either need to put the sensors and the communications equipment onto the vehicle itself, which makes each individual vehicle fully automated; or you put some sensors and equipment on the truck, but a lot more into the road and into roadside infrastructure. For some analysts this raises questions of whether these trucks can be truly called ‘autonomous’, but for drivers that’s a side issue – whether the kit is in the truck or in the road, it will operate without a driver. Another key issue for driverless is the existence of secure and very quick digital communications between the truck, its sensors and off-board equipment. This is why the development of 5G technology is so vital to proponents of the technology. And why – in the light of criminal gangs and the way state-to-state conflict is taking place online – security and hacking remain crucial questions that demand that a driver always be in control of the vehicle.
More importantly, these issues mean that driverless requires substantial infrastructure investments, in communications technology and in the deployment of sensors around our roads. But who will pay for these investments? The UK’s investment in Smart Motorways in an indication that state investment is likely to be needed. Some argue that digital 3D maps of entire countries will need to be developed, and it is true that some companies are developing this technology. But the actual mapping of most countries would be costly, and who would pay for this? Many countries may balk at the idea of giving a foreign company the opportunity to build a 3D map of all their cities and public spaces. Such technology would have huge military value, because it would allow the construction of digital models of entire cities that troops could train on. It would also assist in the accurate targeting of guided munitions.
But there are other obstacles that the technology must overcome before it can be deployed in substantial numbers. Regulations on international traffic are specific about the need for a driver in a road vehicle for example. International standards need to be developed for the testing of sensors and software as well. Liability issues need to be clarified in case of a crash: who would be responsible, the makers of the sensors, the makers of the software, the providers of the communication equipment, or the makers of the vehicle? Or all of them? And what about the Ethical issue embodied in the ‘trolley problem’ (the ethical dilemma faced by the driver of a vehicle who must choose to kill either one person tied to the track or several people)? Can we allow software to have life and death decision-making power? Here too efforts need to be made to develop ethical guidelines for software developers. Work is ongoing in all these areas, but it will probably be years before all of these issues have been ironed out.
And we have no idea how the public will take to driverless technology. The safety claims around driverless are unlikely to materialise give that the technology will coexist with human operated vehicles for a substantial time, and this will make accidents happen. Poor maintenance, solar flares, freak weather, power outages, hacking and vandalism by criminals or teenage delinquents, may all result in lethal accidents that stoke public scepticism.
The environmental claims are also questionable. While automated vehicles may use fuel more efficiently, and could allow reduced fleet sizes, the environmental cost of the technology they carry needs to be factored in, as well as the climate cost of investing in new factories and new infrastructure around the world. Furthermore, the predicted reduction in fleet sizes seems unlikely in an environment of competition. It seems that if we take a holistic view, the introduction of driverless may come at an increased environmental cost.
And in the meantime, drivers are having every aspect of their work analysed by sensors and telematic technology, their physiology analysed by biometric sensors, their communications analysed by AI, and their routes and customer interactions recorded, criticised and eventually determined by an algorithm. Sensors record when they open the door, when they plug in their seatbelts, when they start the engine, and so on. They are then being benchmarked against each other and put under increased pressure to operate within machine-decided parameters. This is creating a much higher-pressure environment for drivers, and increases their vulnerability. The use of sensors to measure wear and tear on equipment and engines could result in improvements, or it could be used to push these items to the limits of their endurance. Without drivers in the vehicle who will maintain the sensors? Who will have an incentive to ensure that the trucks are safe?
This use of technology should become a key issue for all workers in the road transport and logistics sectors, because it undermines their work processes.